When did Windows start accepting forward slash as a path separator?

[Stephen Horne]

The fun you can have on windows, I've
managed to create files that cannot be removed

I've had that one simply by saving a web page in IE. Seriously
annoying.

 

[Stephen Horne]

Stephen Ferg:

It was? My copy of Norton's Guide to the IBM PC makes
the explicit statement that directories were directly influenced by
unix. I actually regard this as a point of similarity. The character
used is a minor issue.

Directly influenced, yes - but as DOS 1 didn't have subdirectories but
did have options marked with '/', the unix subdirectories couldn't be
duplicated exactly in DOS 2. The character used *is* a minor issue,
but it is still a difference - what Stephen Ferg wrote is, I believe,
valid.

Even if DOS was internally happy to accept '/' as a path separator,
still the 'accepted as a truism' bit rings true to me - what is
accepted as an obvious truth is not always actually true, and to most
people users DOS was the command line and Windows was just a shell on
top of DOS - the 'truism' was what people could see, not what was
hidden away in the INT 21h calls that only programmers cared about.

 

[Bengt Richter]

I've had that one simply by saving a web page in IE. Seriously
annoying.

Do you mean windows exploder will not even delete it with right-click, delete?
Or doesn't it show up at all? (how do you know it exists? dos dir?)

Regards,
Bengt Richter

 

[Andrew Dalke]

Stephen Feng:
Me, arguing that the directory structure between DOS and its Windows
derivatives is a point of similarity, compared to the other differences.

Stephen Horne:

Directly influenced, yes - but as DOS 1 didn't have subdirectories ...

Right. I said that as well ... Plus, Stephen's Feng's comparison
is predicated on comparing how directories are accessed, so it's
cheating to go back to when DOS didn't have directories.

what is accepted as an obvious truth is not always actually true,

A truer word has rarely been said.

BTW, the similarity-but-not-the-same between DOS and unix
caused me problems when I started using unix. See, under DOS
there's this great utility called "unerase" which can be used to
recover files accidentally deleted. Only I would deliberately
delete a file in order to probe the system; eg, is this file really
used?

And I wanted to see if the script really used the C compiler....

Had to reinstall IRIX to fix that one.

Andrew
(e-mail address removed)

 

[Cousin Stanley]

The fun you can have on windows, I've

I've had that one simply by saving a web page in IE. Seriously
annoying.

| Do you mean windows exploder will not even delete it with right-click, delete?
| Or doesn't it show up at all? (how do you know it exists? dos dir?)

Under Win98 they show up
AND can NOT be deleted from the Right-Click context menu ...

They can be deleted with no problem under DOS ....

The problem in Win98 for me came from saving web pages
with extra LONG file names ....

 

[Stephen Horne]

Do you mean windows exploder will not even delete it with right-click, delete?
Or doesn't it show up at all? (how do you know it exists? dos dir?)

There was no way I could find to delete it - not even writing a C++
program directly calling the DeleteFile API call.

I don't remember exact details - not even for sure whether it was
Win98 or Win2000 (it was too long ago for WinXP). I do remember that
the non-deletable file was in a deeply nested set of pointless folders
created by saving that web site.

Basically, it ended up sitting on my machine annoying me for ages -
I'm not sure if a new machine or a reinstall put an end to it (it was
my work machine, which tends to get fewer reinstalls).

 

[Grant Edwards]

Internally, the routines in the API have accepted '/' as a path
separator for a long, long time. It's the command-line
processors that kick up a fuss.

And that could always be fixed by running a real shell a-la MKS,
Cygwin, etc.

 

[JanC]

I do remember that the non-deletable file was in a deeply nested set
of pointless folders created by saving that web site.

IIRC that is or was a bug on (some) Windows versions: the path+filename you
have to pass to the Delete File API function could be too long for the
reserved string buffer length (256 bytes), or something like that...

 

[Christos TZOTZIOY Georgiou]

Do you mean windows exploder will not even delete it with right-click, delete?
Or doesn't it show up at all? (how do you know it exists? dos dir?)

I have 110MB (or so they seem) worth of files (2893 files) in my W2k SP4
laptop's c:\recycler\nprotect folder that can't be accessed or deleted.
Norton Protect is long uninstalled, but operating on the files gives
(example):

C:\RECYCLER\NPROTECT\00041753.
The system cannot find the file specified.

The disk shows as healthy in Disk Manager, and scandisk or
what's-its-name runs successfully. Go figure. Ils sont fous ces
Redmontains! (index finger running wild circles)

 

[Christos TZOTZIOY Georgiou]

BTW, the similarity-but-not-the-same between DOS and unix
caused me problems when I started using unix. See, under DOS
there's this great utility called "unerase" which can be used to
recover files accidentally deleted. Only I would deliberately
delete a file in order to probe the system; eg, is this file really
used?

Safe wisdom: mv (move, ren for old DOS) would be better, right?

And I wanted to see if the script really used the C compiler....

Had to reinstall IRIX to fix that one.

If 'inst same' wasn't enough, then that should have been a really old
version of Irix...

 

[Andrew Dalke]

Christos TZOTZIOY Georgiou:

Safe wisdom: mv (move, ren for old DOS) would be better, right?

Sure. I know that *now*.

If 'inst same' wasn't enough, then that should have been a really old
version of Irix...
--

A Personal Iris running 3.0.4, in 1990

Andrew
(e-mail address removed)